Reliability of self-reported diagnostic radiation history in BRCA1/2 mutation carriers

We assessed reliability of self-reported diagnostic radiation history in BRCA1/2 mutation carriers with and without breast cancer. Within the frame-work of the HEBON study, 401 BRCA1/2 mutation carriers completed a baseline (1999-2004) and a follow-up questionnaire (2006-2007). Test-retest reliability of self-reported exposure to chest X-rays, fluoroscopies and mammograms was assessed for the entire study population and by case status. Overall proportion agreement on reporting ever/never exposure was good (> 75%), while the corresponding kappa coefficients were between 0.40 and 0.75, indicating at least moderate reliability beyond chance. Reliability of number of exposures was also good (> 75%). Proportion agreement on reporting age at first mammogram was low (40%) for exact consistency and moderate (60%) for consistency ± 1 year. Reliability of age at first mammogram was higher for cases than for unaffected carriers (P < 0.001) but this difference disappeared when excluding diagnostic mammograms (P = 0.60). In unaffected carriers proportion agreement on age at last mammogram was 50%. In general, the direction of disagreement on all items was equally distributed. More consistent reporting was mainly determined by a younger age at questionnaire completion. In conclusion, inconsistent self-report of diagnostic radiation by BRCA1/2 mutation carriers was mainly non-differential by disease status.